NO. 9 FORECAST OF U. S. PRECIPITATION — ABBOT 6j 



their magnitudes would be greater in proportion as the mean de- 

 parture of his place bears to 20 percent. 



As actual cases, farmers living near Albany, Oreg., or Augusta, Ga., 

 both by table 14 lying in the 26-percent class of table 15, may expect, 

 according to table 15, during the 9 years 1959-1967, the numbers of 

 4-month averages found in table 14 to differ as follows from the 27 

 mean 4-month departures from normal precipitation they will actually 

 experience : Four cases less than 7 percent ; four cases between 7 and 



13 percent; six cases between 14 and 26 percent; six cases between 26 

 and 39 percent ; three cases between 39 and 52 percent ; and four cases 

 over 52 percent. Farmers living near one of the cities of the 20-per- 

 cent class might expect this same division of the 27 cases for the 

 4-month mean departures from normal precipitation, but these de- 



partures would be smaller in percentages in the ratio — . It will be 



26 



for their judgment to dictate whether it is worth while to procure 

 from the Smithsonian Institution, and make use of this paper, "A 

 Long-range Forecast of U. S. Precipitation." 



COUNTRY-WIDE TRENDS IN PRECIPITATION 



The maps of the United States presented below show large areas 

 over which similar forecasts prevail. This should be helpful to in- 

 terested persons who reside at a distance from the 32 cities for which 

 forecasts were made. 



I have been interested to search further to see if similar trends of 

 precipitation sometimes prevail over the whole United States. Table 



14 gives the actual departures of 3-month consecutive means of pre- 

 cipitation as averaged over three 4-month intervals per year, 1950- 

 1958. A working table of these results was prepared, giving the 32 

 departures from normal of the cities employed in each line of a table 

 of 27 lines, 3 lines per year for 9 years. Recording separately plus 

 and minus departures, sums were taken for each line. These plus and 

 minus departure-sums were plotted in figure 12, lower two curves. 

 Plus sums are given in full lines, minus sums in dotted lines. 



The plus and minus departure curves run generally in opposite 

 directions, and in some 4-month intervals are widely separated. In 

 such cases of wide separation the 4-month intervals were strongly 

 heavy in precipitation if the high points are on full lines, and strongly 

 drought-prevailing if dotted. With this explanation it is seen that the 

 autumn of 195 1 and winter of 1952 were wet periods generally for 

 the whole United States, and similarly from the summer of 1957 



